

For example, the formula-induced pyridoxine deficiency seizures, discussed later, occurred before infants' pyridoxine requirement was known. It is interesting that most of the errors relate to historical developments in medicine and might not have happened in another era.

But that does not mean the errors were unavoidable. I use the word error to indicate an act that unintentionally deviates from what is correct. Indeed, one of his stories tells of an epidemic of Pseudomonas infection in premature infants that resulted from contamination of delivery room resuscitation equipment at Magee-Women's Hospital in Pittsburgh. Stories about the mistakes in neonatology are both educational and fascinating, reminiscent of the wonderful medical detective stories of Berton Roueché. And I worry also about the safety of currently accepted, but untested, therapies that I have used.

But many colleagues still surprise me by not being more cautious about new therapies. Silverman, 1 in his 1980 book Retrolental Fibroplasia: A Modern Parable, discusses examples of therapies that led to "disasters" (Chapter 10, Medical Inflation). A classic example is using high oxygen concentrations in caring for premature infants, resulting in retrolental fibroplasia (RLF). This lack of controlled measurements has certainly been true in neonatology. In any rapidly changing medical field, treatments and procedures may be instituted without controlled outcome measurement that might reveal untoward effects.
